(aperçu de) la physique des oxydesde cobalt et de sodium Na xCoO2

Marc-Henri JulienLaboratoire de Spectrométrie Physique – UMR 5588

Université J. Fourier – Grenoble I

Cédric de Vaulx (PhD 2004-2006) NMR : H. Mayaffre, C. Berthier, M. Horvatic (Grenoble)

Material synthesis & crystal growthP. Lejay, P. Strobel, H. Muguerra (Grenoble)C.T. Lin & D.P. Chen (MPI Stuttgart),J. Wooldridge, G. Balakrishnan & M.R. Lees (Coventry)J.M. Tarascon, M. Armand (Amiens)V. Pralong (Caen)

XRD & SQUID: P. Bordet, V. Simonet (Grenoble),S. Hébert (Caen)

Acknowledgements

Objectifs de la présentation

1) Aperçu global du diagramme de phase de Na xCoO2et de ses propriétés électroniques principales(corrélations électroniques)

2) Particularité majeure de ces composés :mise en ordre des Na + etimpact sur les propriétés électroniques

Cu-basedHTc cuprates

Sr2RuO4- metallic, multibands- p-wave superconductor

Nickelates La2-xSrxNiO4- insulating- Spin and charge stripes

Cu-based spin-ladders- spin gap- d-wave superconductor

Cu-based delafossites(LaCuO2.66, etc.)- No superconductivity- Insulating- Magnetic LRO

Change Cu for RuChange Cu for Ni

Cut CuO2 planesinto slices Change for

NaxCoO2

Search for triangular analogs of high T C’s

AxMO2 genealogy

AxCoO2:A = Na, Li , K,

Sr, Ca, etc.

Intercalation materials, ionic conductors

LiNiO2, NaNiO2, AgNiO2, magnetism

LiVO2, LiMn02, cathode materials

« Li-ion »

Cobaltate genealogy

« misfits »

Na1CoO2 a band insulator ?

Na+

(O2-) x 2 Co3+

eg

t2g(e’g & a1g bands)

OctahedralCF splitting~ 2 eV

No spin shift(susceptibility)

Extremelyslow nuclearRelaxation(> 10 s)

NMRde Vaulx et al., PRL 2005Lang et al., PRB 2005

Na1CoO2 is a band insulator (Co 3+, S=0)

0.00.2

2.02.53.0

0 100 200 30010-2

10-1

100

101

Kc [

%]

T1-1

[s-1]

T[K]

23Na

23Na

59Co

59Co

de Vaulx et al ., PRL 95, 186405 (2005)

Korb

S = 0 N(EF) = 0

Although obtaining pure x=1 compounds can be challenging

LiCoO2 : Ménétrier et al., Electrochem. Solid-State Lett., 11, A179 (2008) AgCoO2 : Muguerra et al., J. Solid State Chem. 181, 2883 (2008)

A1CoO2 are all band insulators

0 100 200 3000.0

0.5

1.0

1.5

2.0

χ (

10 -5

em

u m

ol-1 O

e-1)

T (K)

AgCoO 2

χχχχ = χχχχorb + χχχχimp= const + 1/T

H. Muguerra et al.

(A=Li, Na, K, Ag, etc.)

χχχχorb

Na+

e-

Hole-doping the x=1 band insulator

Co3+ (S=0) Co4+ (S=1/2)

eg

t2g

By removing x (Na+) from the layers

(e’g & a1g bands)

Orbital and band structure issues

Co4+

Co3+ S=1Not discussed here…

Co3+ intermediate spin state ?(local symmetry breaking)No spin-state transition ever observed in Na xCoO2

a1g & e’g bands,fermiology…

Hybridization with O 2p orbitals Orbital ordering, spin-orbit coupling ?

Khaliullin and coworkers

predominant a1g characterEF

D.J. Singh, PRB 2000

Two kinds of Na sites

Na2 Na1Co lattice

M. Roger et al., LT25 proceedings & Nature 445, 631 (2007)

Observed phases depend onsynthesis method & conditionselectrochemical, chemical, direct,ion-exchange, pulsed laser deposition, etc.

Accurate measure of x(Na)is challenging (while physicalproperties can change withvariation of x ~ 1%)

Surface ordering differentfrom bulk

Sample aging (Na losses)

Good samples are clean!(sharp NMR lines, Shubnikov–de Haas oscillations)

Lots of material issues

0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00

0

15

30 (b)

x(Na)

(1)(2)

(2)(1)

(5)

Sam

ples

(1+3)

(c)

00 (?)0

20 K27 K

22 K

TM (

K)

Finite number of stable phasesUbiquitous phase separation for 0.67 < x < 0.9

NaxCoO2Delmas, Solid State Ionics (1981)de Vaulx, PRL (2005)Lee, Nature Materials (2006)Li xCoO2Van der Ven, PRB (1998) Ménétrier, J. Mater. Chem. (1999)Marianetti, Nature Materials (2004)

MH Julien et al., unpublished

(some of the) stable phases

T and x dependent Na + mobility

0 100 200 300 400 500

ρρρρ (mΩΩΩΩ cm)

Freezing180 K

High T anomaliesin transportmeasurements

Slow dynamicsdetected in NMRdown to ~180 KJulien et al., PRL 2008Lang et al., PRB 2008

Blangero et al., PRB 2008

x=0.63

x=0.75

T (K)

Impact of Na + ordering on magnetism

Schulze et al., PRL 100, 026407 (2008)Na0.8CoO2

T(K)

Apparition of a magnetic transition at 8 K,depending on cooling protocol above ~200 K

Neither 1 nor 2 59Co NMR sites

Mukhameshin, PRL (2005) Alloul et al., EPL (2008)Julien et al., PRL (2008)

Na0.75CoO2

3 different sites (at long time scale, ~ 10 -6 s)30% (not 75%) of localized Co3+

no Co4+

« charge disproportionation », « charge ordering »,« electronic texture », electron localization

But itinerant character remains!

Finite number of 23Na NMR linesDemonstrates spatial ordering of the different Co states

Na1

Hyperfine field transferredfrom Co to Na

23Na spectrum ifCo states were atrandom positions(simulation)

Julien et al., PRL (2008) x=0.75

Co3+

Co3.3+

Co3.4+

NMR summary

(possibly)

59Co NMR 23Na NMR

One example ofpossible

ordered pattern

Na+ ordering is observed

M. Roger et al., Nature 445, 631 (2007)

e.g. vacancy clustering for x ~ 0.8

Na1-Na2 stripe correlations for x < 0.7 (G. Collin)

Details are controversial (accuracy of x values ?)

M. Roger et al., Nature 2007

Na1 should dislikeCo4+ more than Co 3+

(Coulomb repulsion)

Electrostatic potential from Na+ ionsModulates charge density in cobalt planes

Na

Co

Na

« Texture » in cobalt planes depends on 3D stackingof Na+ ordering pattern

Relevance to high Tc cupratesDopant–induced electronic inhomogeneity

High values of the thermoeletric power

Wang et al., Nature (2003)

Lee et al, Nature Materials (2006)

Magnetic field dependent Strongest at the highest x(Na)

Magnetic properties / electron correlations are involve d

Curie-Weiss bulk magnetization

Magnetic transitionsEx: TM = 22 K for x = 0.75None for x = 0.67 and x = 0.71

A-type antiferromagnetBayrakci PRL (2005), Helme PRL (2005)FM planes no frustration too bad…

Magnetic properties

Signature of electron correlations

ARPESHB Yang, PRL (2005)

As well as in transport measurements(T-linear resistivity, m*/m ~ 3 - 10)

U ~ 2 – 4 eV

Na0.5CoO2

Huang, PRB 2004

Transitionmétal-isolant

53 K

TransitionMagnétique

88 K

Fermi surface instability at x=0.5Very small charge differentiation

G. Lang et al., PRB 2008 J. Bobroff et al., PRL 2006

Co3+/Co4+ mixtureNa order FS reconstruction Nesting Transition

Weak electron correlation at x = 0.3

Foo et al., PRL (2004)

Pauli-like χχχχspin

ρρρρ ∝∝∝∝ T2 below ~ 30 K

Fermi liquid ?

Takada et al., Nature (2003)

Na

CoO2

H2O

Water-induced superconductivity

Schaak et al., Nature (2003)

Tc max = 4.8 K

in Na0.3CoO2yH2O

Unconventional superconductivity ?

Type II Short coherence length (~100 Angstrom) Singlet pairing Pairing symmetry ?

AFM spin fluctuations and proximity to magnetic orderi ng

Ihara et al ., JPSJ (2006)

Full electrochemicalextraction of Li +

Amatucci, Klein & Tarascon,J. Electrochem. (1996)

LiCoO 2 CoO2

The x = 0 limit CoO2 is not a Mott insulatorBut a correlated metal

59Co NMR :de Vaulx et al., PRL 2007

1 10 1005

10

15

20

(T1T

)-1 [s

-1K

-1]

T [K]

T* ≈ 7 K

AF (q≠≠≠≠0) spinfluctuations

Fermiliquid

Korringa

Curie-Weiss

(Metal-insulator) Mott transition occurs forU/t ≈≈≈≈ 10-12 on the triangular lattice

NaxCoO2 summary Carrier concentration can be varied overa wide range (1 elect./cell)

Electron correlationsthroughout the phasediagram

Unexpectedly, correlationsare strongest at high Naconcentrations

Correlations can beenhanced by Na + potential

Limited analogy with high Tc’s

AFM FM

G. Lang et al. PRB (2008)

Correlations ++Impact of Na ions

ThermopowerMagnetism

Somecorrelations